This repository includes:
- A 1,037 user dataset with students and non-students manually labeled.
- Six final student identification models with varying accuracies, f1 scores and coverages for unique applications.
- Project code used to create and analyze the student identification models.